A Name Below the Picture

J.-L. Moulin spent about two and a half months preparing a portrait that could be woven in silk. His employer, Didier Petit, would not allow his name on the finished picture. The historian Maxime Dehan, who has worked with the Petit family's papers, records the refusal. The name of the manufacturing house appeared beneath the image instead.

This was Lyon in 1839, and the firm needed something to exhibit in Paris. Its usual business included furnishing fabrics and church textiles. Moulin proposed a portrait of Joseph-Marie Jacquard, whose mechanism had changed the work of figured weaving. A likeness of the inventor, produced through his own device, would give the manufacturer something more particular to display than another patterned length of silk.

The result looks like a fine engraving. Jacquard sits in a padded chair, holding a pair of compasses. A small loom mechanism and some punched cards occupy the table beside him. Tools hang behind his shoulder. Light passes through the thin curtain over a window with a broken pane. There is no ink making the shadows: the image is woven into the cloth.

Before the cardboard

The painting came first. In the aftermath of the Lyon silk workers' revolt of 1831, the city commissioned Claude Bonnefond, director of its school of fine arts, to paint Jacquard. The inventor was still alive when Bonnefond drew him. The portrait reached the Paris Salon in 1834, the year of his death.

By choosing Jacquard for its exhibition piece five years later, Didier Petit et Cie was taking an already recognisable civic image into the trade it celebrated. The work was entrusted to Michel-Marie Carquillat, a weaver who became a specialist in pictures made from silk. Moulin's part came between Bonnefond's likeness and Carquillat's loom.

A metteur en carte turned a picture into a weaving design on squared paper. The squares gave a way to relate positions in the image to threads running lengthways through the loom and successive passages of thread across them. Curves and shaded surfaces had to become choices about which threads would show and which would be concealed. Moulin worked with help from Joseph-Victor Vibert, a professor of engraving, according to Dehan's account.

Black thread against a white ground could suggest the marks of an engraving. Different interlacings could expose more or less of the dark patterning thread, giving the eye a face, a fold of cloth or a translucent curtain. This was not a matter of supplying a grey thread for every shade Bonnefond had painted. The structure of the weave had to carry the modelling.

The stiff cards had to be punched from this paper design and laced into a sequence for the machine. Preparing the image and preparing the cardboard were distinct stages of the work.

Which threads rise?

The Jacquard apparatus sat above the loom, connected to the warp by a multitude of cords.

The warp consists of the threads held lengthways under tension. To weave, some must rise while others stay down, opening a passage for the shuttle carrying the crosswise thread, or weft. In a complex figured fabric, the choice changes repeatedly. The Jacquard mechanism made that selection from a prepared sequence of cards.

A rotating carrier brought each card against a bank of horizontal needles. Where the card had a hole, the corresponding needle could enter. Where it had no hole, the solid cardboard pushed the needle backwards. Each needle governed the position of an upright hook.

The hook's position decided whether its thread would rise.

A hook left in line could be caught by a rising frame and lifted, pulling on its harness cords and raising the associated warp threads. A needle pushed back by the card moved its hook clear of the frame, leaving those threads down. The shuttle could then cross the opening. With the next card came a new selection.

Posselt's later manual also explains how this apparently neat decision could fail. Cards needed enough stiffness to resist the needles, and their locating holes had to fit the carrier's pegs. Damp could warp them. A badly aligned carrier could make the needles strike the edges of the intended holes and produce the wrong lift. One diagnostic was to mark the needle tips with paint or grease, let them meet a card, and inspect where the marks landed.

The weaver still had to pass the shuttle, bring each new weft against the growing cloth and maintain the regularity on which a smooth shaded surface depended. A correct sequence of lifts could not compensate for uneven weaving.

Ada Lovelace supplied an early measure of the portrait's preparation. In her notes on Charles Babbage's proposed Analytical Engine, published in 1843, she reported that the woven Jacquard portrait had required 24,000 cards. She used the example in a discussion of how repeating cycles of operations could reduce the number of cards a calculating engine needed.

A closer look · 3:24 · English on-screen text

From a drawing to the lifting threads

Play film

A portrait for the Queen

Charles Babbage watched the design being woven during a visit to Lyon on his way to Turin in 1840. He already owned a copy, given by a friend, but wanted another to take to the Grand Duke of Tuscany in Florence. In his 1864 memoir, he remembered spending hours watching the loom. He also recorded that the pictures were not ordinarily made for sale; he had been allowed to buy one as a favour.

In Turin he changed his plans for the gift. He asked whether he might present the silk picture to the Queen of Sardinia. This required the King's permission, and an audience was arranged.

Babbage brought a large portfolio. He unpacked the portrait from sheets of exceptionally light protective paper, and the King inspected it on both sides and asked about its construction. Repacking was less orderly. In Babbage's account the sheets escaped into the air, and the King took the textile from him while he tried to catch the paper. Eventually both men were on their knees gathering it up.

The framed copy in Babbage's London drawing room offered a different kind of difficulty. When he showed it to the painter David Wilkie and asked what it was, Wilkie identified an engraving. Babbage brought him closer. After considering another kind of engraving, the painter eventually wondered whether it could be lithographyLithography×A printing method using a flat stone or metal plate. The image attracts ink while the blank areas repel it.Merriam-Webster · lithography ↗. That, too, is Babbage's recollection; the silk's convincing imitation of printed lines was the point of his test.

In 1842, before showing Prince Albert and the Duke of Wellington his Difference Engine, Babbage stopped them in front of the portrait. He thought it would help explain calculating machinery. Wellington took it for an engraving; Albert recognised it. Babbage had adopted the Jacquard principle for his Analytical Engine before this portrait existed. The cloth gave him a finished object through which to explain a sequence of mechanical instructions.

The weaver presents his work

The portrait travelled in another form, too: as a picture inside a picture. On 24 August 1841, the duc d'Aumale, a son of King Louis-Philippe of France, visited Carquillat's workshop in Lyon's Croix-Rousse district. A woven commemoration followed in 1844.

Bonnefond designed the scene. Carquillat stands before the uniformed visitor, presenting a copy of the Jacquard portrait. Behind them a large loom fills the workshop, its card chain hanging above the group. A woman and child occupy the right side. This is a composed exhibition picture made after the visit, rather than an unposed record of the afternoon.

Credit was not identical on every surviving Jacquard portrait. The Library of Congress copy reproduced here names Carquillat at its foot, as does the Metropolitan Museum's example. Moulin's name is absent. Dehan records that he later made a copy of the weaving design, which he gave to Lyon's textile museum in 1895.

On the 1844 workshop picture, the division of work is explicit. Beneath the scene, separate inscriptions name Bonnefond as painter, Carquillat as weaver and A. Manin as the man who drew and prepared the mise en carte. Below the title and the visitors' names is a dedication to the King, signed by Carquillat.